In the Original Contribution entitled "Enhancement of Clinicians' Diagnostic
Reasoning by Computer-Based Consultation: A Multisite Study of 2 Systems"
published in the November 17, 1999, issue of THE JOURNAL (1999;282:1851-1856),
there were discrepancies in a table. On page 1854, the following table replaces Table 1.
Table 1. Decision Support Consultation
Effects Based on Presence of the Correct Diagnosis*
No. of Cases
No. (%) of Consultations
Correct Diagnosis
Positive†
Negative‡
Net Gain§
Without Decision Support System
With Decision Support System
Experience Level
Students
647
168 (26.0)
227 (35.1)
96 (14.8)
37 (5.7)
59 (9.1)
Residents
648
283 (43.7)
319 (49.2)
70 (10.8)
34 (5.2)
36 (5.6)
Faculty
639
314 (49.1)
335 (52.4)
66 (10.3)
45 (7.0)
21 (3.3)
Decision Support System
ILIAD
972
390 (40.1)
431 (44.3)
101 (10.4)
60 (6.2)
41 (4.2)
QMR
962
375 (39.0)
450 (46.8)
131 (13.6)
56 (5.8)
75 (7.8)
All cases
1934
765 (39.6)
881 (45.6)
232 (12.0)
116 (6.0)
116 (6.0)
*Within each row of the table, percentages are based on the total number
of cases in that row. QMR indicates Quick Medical Reference.
†A consultation was deemed positive when clinicians generated
the correct diagnosis as part of their differential diagnosis after a decision
support system consultation but did not include it on their initial preconsultation
diagnosis list.
‡A consultation was deemed negative when clinicians included
the correct diagnosis in their initial preconsultation diagnosis list but
excluded it from their postconsultation diagnosis list after using a decision
support system.
§Values represent the difference between the number of correct
cases after a decision support system consultation and the number correct
before consultation.